Koffei

Daily CyberNews · 05/10/2026

‹ 02/10/202606/10/2026 ›
Today in 30 seconds
  • 🔴 Citrix NetScaler SAML zero-day CVE-2026-88779 is exploited and in KEV; patch by October 7.
  • 🔴 FortiMail CVE-2026-104286 is still unpatched and exploited; disable IBE or restrict management access.
  • 🔴 Warlock ransomware hits unpatched SharePoint ToolShell servers; patch SharePoint and hunt for VS Code tunnels.
  • 🟠 Self-managed GitLab: update to 19.2.4, 19.3.2 or 19.4.1 for AI Gateway CVE-2026-90970.
  • 🔴 TA419 phishes US AI policy experts via AitM pages; move them to passkeys.

Vulnerabilities 5

🔴 Exploited or active

Citrix NetScaler SAML zero-day knocks deployments offline

CVE-2026-88779CVSS 8.7NetScaler SAMLin KEV

WhatA memory-overflow flaw in NetScaler ADC/Gateway configured as SAML SP or IdP is under targeted exploitation and can cause denial of service.

DoPatch NetScaler instances using SAML; the CISA federal deadline is October 7.

Details and sources
CVE-2026-88779 (CVSS 8.7) is a memory-overflow flaw in NetScaler ADC/Gateway instances configured as a SAML service provider or identity provider, under active targeted exploitation that can trigger denial-of-service. Citrix credited Bishop Fox and watchTowr; CISA added it to the Known Exploited Vulnerabilities catalog with a federal patch deadline of October 7. The Hacker News (source)
🔴 Exploited or active

FortiMail critical zero-day actively exploited, no patch yet

CVE-2026-104286CVSS 9.8FortiMailno patch

WhatA path traversal and null byte flaw in the FortiMail management interface lets unauthenticated attackers write arbitrary files, with fixes 7.4.9, 7.6.7 and 8.0.2 still pending.

DoDisable IBE or restrict management-interface access until Fortinet releases fixed versions.

Details and sources
CVE-2026-104286 (CVSS 9.8) is a path traversal / null byte injection flaw in the FortiMail management interface letting an unauthenticated attacker write arbitrary files on the underlying system. Affects FortiMail 7.2–8.0.1; Fortinet has no fix out yet (7.4.9/7.6.7/8.0.2 are pending) and recommends disabling IBE or restricting management interface access. CISA gave federal agencies until October 4 to triage and mitigate. Fortinet · BleepingComputer (source)
🟠 High impact

GitLab patches critical 9.9 AI Gateway sandbox-escape flaw

CVE-2026-90970CVSS 9.9GitLab AI Gateway

WhatA logged-in user with Duo Agent Platform access could escape the prompt-template sandbox and possibly run commands on self-hosted AI Gateway instances.

DoUpdate self-managed GitLab to 19.2.4, 19.3.2 or 19.4.1.

Details and sources
CVE-2026-90970 let a logged-in user with Duo Agent Platform access escape the prompt-template sandbox via a crafted flow configuration, potentially enabling arbitrary command execution on self-hosted AI Gateway instances. GitLab-hosted/managed gateways were already patched; self-managed customers should update to 19.2.4, 19.3.2, or 19.4.1. GitLab · The Hacker News (source)
🔴 Exploited or active

Cisco Catalyst SD-WAN Manager auth-bypass added to CISA KEV

CVE-2026-76504CVSS 9.8Cisco SD-WANin KEV

WhatA URI-handling flaw lets unauthenticated attackers reach the SD-WAN Manager admin API via a crafted /j_security_check request, exploited since September 2026.

DoPatch Cisco Catalyst SD-WAN Manager now; the federal deadline was October 3.

Details and sources
CVE-2026-76504 (CVSS 9.8) is a hex-encoding/URI-handling flaw letting unauthenticated remote attackers reach the admin API by sending a crafted request to /j_security_check. Confirmed exploited in the wild since September 2026; federal agencies had until October 3 to patch. Cisco · CISA · The Hacker News (source)
🟡 Worth knowing

Apple CoreGraphics PoC surfaces, WhatsApp delivery path unconfirmed

CVE-2026-86950Apple CoreGraphicsPoC crash

WhatA public PoC for the CoreGraphics out-of-bounds write shows a crash, not code execution, and a claimed WhatsApp delivery path was retracted.

DoMake sure Apple devices have the September 28 update installed.

Details and sources
CVE-2026-86950, patched by Apple on September 28, causes a controlled out-of-bounds write in CoreGraphics; a public PoC (Sept 30) demonstrates a crash, not code execution. Researchers flagged new PDF font-scanning code in recent WhatsApp builds (26.37.73–26.38.74), but the write-up does not actually describe or test a WhatsApp delivery path, and the speculation was retracted shortly after publication. No vendor advisory confirms in-the-wild use. Apple · The Hacker News (source)

Breaches / Incidents 3

🟠 High impact

Danish university DTU breach exposes up to 200,000 people

DTUDenmark~200,000 peoplestolen credentials

WhatAttackers used stolen credentials to access DTU's identity system, possibly exposing CPR numbers, addresses and next-of-kin details for about 200,000 current and former users.

Details and sources
Attackers used stolen credentials to access the Technical University of Denmark's identity and access management system (DTUBasen), potentially exposing two decades of records — Danish CPR numbers, names, home addresses, profile pictures, work details, and next-of-kin contacts for roughly 40,000 active and 160,000 former users. DTU says it cannot determine exactly what was downloaded. BleepingComputer (source)
🟡 Worth knowing

MetaMask discloses security incident, exits affected Ethereum validators

MetaMaskstaking validatorsincident

WhatMetaMask reported an ongoing incident in part of its infrastructure and is exiting affected staking validators, with no threat to wallets or funds identified.

Details and sources
MetaMask reported an ongoing security incident affecting part of its infrastructure and is precautionarily exiting affected validators in its non-custodial staking operations, expecting the exit process to finish by October 7. No threat to wallets or customer funds has been identified, and no cause has been disclosed. MetaMask · The Hacker News (source)
🟡 Worth knowing

ShinyHunters suspect "Rey" reportedly detained in Jordan

ShinyHuntersReyJordanFBI

WhatA suspected ShinyHunters member known as Rey has reportedly been detained in Jordan and is cooperating with the FBI.

Details and sources
A suspected member of the ShinyHunters extortion group, going by "Rey," has reportedly been detained in Jordan and is said to be cooperating with the FBI to help identify other group members. BleepingComputer (source)

Malware / Ransomware 2

🔴 Exploited or active

Warlock ransomware exploits SharePoint ToolShell flaws to disable security tools

WarlockSharePoint ToolShellCVE-2025-1055

WhatWarlock hit at least four organizations through unpatched SharePoint ToolShell flaws, killing security tools with a vulnerable driver and using VS Code tunnels before deploying ransomware.

DoPatch SharePoint Server, and hunt for web shells, SYSVOL-staged payloads, vulnerable driver loads and VS Code tunnels.

Details and sources
Warlock (aka Gold Salem, Longlegs, Storm-2603), suspected China-linked, hit at least four organizations in Portuguese- and Spanish-speaking regions — including a water utility, a telecom provider, a regional government body, and a university — via unpatched SharePoint Server "ToolShell" vulnerabilities. The chain deployed web shells to harvest ASP.NET machine keys, used DLL sideloading, pulled payloads from cloud storage, abused a vulnerable driver (CVE-2025-1055) to kill security software, staged ransomware via SYSVOL for mass deployment, and set up VS Code tunnels for remote access before detonating ransomware network-wide. The Hacker News (source)
🟡 Worth knowing

KillSec ransomware operation dismantled, three arrested

KillSectakedownthree arrests

WhatEuropean police arrested three KillSec suspects, including a 16-year-old alleged admin, and seized the leak site, five servers and over 110TB of data.

Details and sources
European police arrested a 16-year-old in Alicante, Spain (suspected main administrator), a man in his 20s in the UK, and a 24-year-old in Romania, and seized KillSec's leak site and five servers along with over 110 terabytes of data, devices, and crypto wallets tied to ransom payments. KillSec began as a hacktivist group in 2021 before pivoting to ransomware/extortion in 2023; investigators link it to roughly 1,000 suspected attacks worldwide, with 500+ confirmed successful. The Hacker News (source)

Threat Actor Activity 1

🔴 Exploited or active

TA419 (China-aligned) phishes U.S. AI policy experts via Microsoft AitM pages

TA419ChinaAI policyAitM OneDrive

WhatTA419 phishes US AI policy experts and defense contractors with impersonation lures leading to browser-in-the-browser OneDrive pages that steal credentials and session cookies.

DoUse phishing-resistant authentication such as passkeys and verify unsolicited outreach before engaging.

Details and sources
TA419, active since at least April 2025 and previously focused on U.S./Japan targets, is running credential-phishing campaigns against AI policy experts at U.S. think tanks, universities, and legal organizations, plus defense contractors and former White House S&T Policy staff. The group impersonates economists and AI policymakers — one lure (Feb 2026) was titled "Request for Feedback on Military Integration of Claude" — and uses reply-triggered shortened-URL redirect chains, Cloudflare Turnstile bypasses, and frameless browser-in-the-browser pages spoofing Microsoft OneDrive logins to run adversary-in-the-middle credential and session-cookie theft while keeping the victim's login appearing successful. Recommended defenses: phishing-resistant auth (passkeys) and verifying unsolicited outreach before engaging. The Hacker News (source)
‹ 02/10/202606/10/2026 ›